Advanced AI services to enhance the baggage count prediction, identify the last baggage, manage misconnects and optimize the entire baggage life cycle

These AI services directly addresses the critical operational challenges by providing highly accurate baggage count predictions, enabling the last baggage prediction and misconnect baggage prediction for transfer flights and comprehensive solution for managing and optimizing the entire baggage life cycle.

The system is built on AI services integrating deep learning and machine learning models with baggage and flight data at ADB airport for accurate baggage count estimation over different time intervals, last baggage identification, misconnect baggage prediction and baggage life cycle management. Key technical features include: • Exploratory data analysis with baggage and flight data by Apache Spark (PySpark) • Apache Kafka-based streaming architecture for real-time data collection from AWS and messaging between services. • AI services for baggage count prediction over different time intervals (daily, hourly, minutely) leveraging by different machine learning (Random Forest, XGBoost, CatBoost, and LightGBM) and deep learning models (CNN, LSTM, CNN-LSTM hybrid, ANN) • Apache Cassandra – NoSQL database for storing high-volume flight & baggage data • Web-based user interface design by React for real-time visualization and monitoring of real-time statistical data and prediction results and RESTful API endpoints for flight, baggage and prediction data • AI service for last baggage prediction and misconnect baggage prediction for connected flights • Microservices-based containerization, independent containerized services for all modules, Linux-based dockerized platform.

• Apache Kafka for real-time data streaming and messaging • Docker containerized deployment on Linux server • RESTful API interface with JSON payload format

Airport operators/authorities and their ecosystem

Through IP licensing and commercial agreements from Siemens